부산대학교 도서관

Hardware Trojans (HT) refer to tiny circuits that adversaries implant for malicious purposes. These circuits operate stealthily and, once triggered, can lead to severe disruptions. Detecting their presence requires the application of suitable test vectors. However, a huge search space often complicates the test generation task. In this regard, evolutionary algorithms are promising candidates due to their ability to explore such search space with effective test vectors. This paper introduces a test generation methodology that effectively utilizes the capabilities inherent in particle swarm optimization (PSO) to successfully attain its intended objectives. The validity of the proposed approach is confirmed by applying it to multiple ISCAS ’85 benchmark circuits. The results show a significant reduction in test generation time, exceeding 50%, alongside an enhancement in test set quality compared to the current state-of-the-art techniques.